On Friday 01 July 2005 05:37 am, Benjamin Niemann wrote:
> [EMAIL PROTECTED] wrote:
>
> > Hi,
> > why is this possible -
> >>>>b = [1,2,3]
> >>>>b[2] = b
> >>>>b
> > [1,2,[...]]
> >>>>b[2]
> > [1,2,[...]]
> >>>>b[2][2][2][2][2]
> > [1,2,[...]]
> >
> > but this is not -
> >>>>x = [1,2,x]
> > Traceback (most recent call last):
> > File "<stdin>", line 1, in ?
> > NameError: name 'x' is not defined
>
> Because the right hand side ('[1,2,x]') is evaluated *before* the value is
> bound to the name 'x' - and at this point there is obviously no name 'x'
> defined.
OTOH, the name "b" was bound to the list, so it can be put into
the list, just like any other name. At that point, the list contains
a series of references, one of which happens to be to itself.
If you are familiar with Unix/Linux filesystems, you've probably
already seen behavior like this with either symbolic or hard links --
it's quite possible to create loops so that a directory is contained
within itself. Such a structure is not strictly a "tree" any more,
but a "directed graph". Those are the data structures terms for
them, and you might try a little googling to learn more about them.
The *really* smart thing is that Python *writes* the list as:
[1,2,[...]]
instead of printing a traceback due to "excessive recursion" which
is what it used to do. This is because the representation method
was changed to catch such circular references and stick in the "[...]"
instead.
--
Terry Hancock ( hancock at anansispaceworks.com )
Anansi Spaceworks http://www.anansispaceworks.com
--
http://mail.python.org/mailman/listinfo/python-list
